that simple survey approaches would be sufficiently good in some cases for
monitoring. They would certainly be an improvement on having no monitoring
information at all, as is currently the case for the majority of the world.
The common feature of these classes of question is that the effects being
investigated are bigger than simply changing the abundance of a few species.
Important differences between habitats or sites are generally determined by the
presence or absence of quite a few species. The same could be said of long-term
monitoring as well. Sophisticated schemes might measure population changes
to within a few percent year on year, and this detail is valuable for analytic
approaches to diagnosis of causes of change. On the other hand, the kinds of
changes that trigger any practical response will rarely be less than a 50% decline.
This was the case for all the most impacted common farmland birds in the
United Kingdom, so that they were listed as national conservation priorities
(Gregory et al. 2002). Changes of this magnitude lead to widespread gaps in
the occurrence of species on scales of a few kilometers or more, and could be
detected by simple relative methods.
Effectively, you get the results (absolute or relative abundance) that you pay
for, in terms of intensity and technical sophistication of effort. Absolute counts,
returning an unbiased estimate of the real number of birds in a specified time
and place, are an ideal. They can be obtained with high effort and technically
appropriate methods. As a result, they are usually reported for a selected range of
species or sometimes only one. The intensity of effort required also means that
absolute results are normally described for relatively restricted areas. The main
cost of technical census approaches comes from the difficulty of removing bias
from the results, for instance by measuring distance to registrations to allow for
differences of detectability across habitats.
Relative abundance data tend to return larger numbers for more abundant
species if, but only if, other things such as habitats or the behavior of species
are similar. They differ from absolute counts in the following important respects.
2 |Bird diversity survey methods
Table 1.1Comparisons of survey and census
Survey CensusPrior knowledge Limited Moderate to good
Area covered Large Restricted
Species covered All or many Selected few or one
Methods Simple Some technicality
Effort Extensive Intensive
Quantification desired Qualitative/relative Quantitative/absolute